Semi-supervised learning for an improved diagnosis of COVID-19 in CT images

• Published in: PloS one 2021-04, Vol.16 (4), p.e0249450-e0249450
• Main Authors: Han, Chang Hee, Kim, Misuk, Kwak, Jin Tae
• Format: Article
• Language: English
• Subjects: Algorithms; Biology and Life Sciences; Computed tomography; Computer-aided medical diagnosis; Coronaviruses; COVID-19; COVID-19 - diagnostic imaging; CT imaging; Datasets; Datasets as Topic; Deep learning; Diagnosis; Diagnostic systems; Humans; Image segmentation; Infections; Lungs; Machine learning; Medical diagnosis; Medical imaging; Medical treatment; Medicine and Health Sciences; Methods; Neural networks; Neural Networks, Computer; Polymerase chain reaction; Research and Analysis Methods; Semi-supervised learning; Signs and symptoms; Supervised Machine Learning; Thorax - diagnostic imaging; Thorax - pathology; Tomography; Tomography, X-Ray Computed; Viral diseases
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0249450

Coronavirus disease 2019 (COVID-19) has been spread out all over the world. Although a real-time reverse-transcription polymerase chain reaction (RT-PCR) test has been used as a primary diagnostic tool for COVID-19, the utility of CT based diagnostic tools have been suggested to improve the diagnostic accuracy and reliability. Herein we propose a semi-supervised deep neural network for an improved detection of COVID-19. The proposed method utilizes CT images in a supervised and unsupervised manner to improve the accuracy and robustness of COVID-19 diagnosis. Both labeled and unlabeled CT images are employed. Labeled CT images are used for supervised leaning. Unlabeled CT images are utilized for unsupervised learning in a way that the feature representations are invariant to perturbations in CT images. To systematically evaluate the proposed method, two COVID-19 CT datasets and three public CT datasets with no COVID-19 CT images are employed. In distinguishing COVID-19 from non-COVID-19 CT images, the proposed method achieves an overall accuracy of 99.83%, sensitivity of 0.9286, specificity of 0.9832, and positive predictive value (PPV) of 0.9192. The results are consistent between the COVID-19 challenge dataset and the public CT datasets. For discriminating between COVID-19 and common pneumonia CT images, the proposed method obtains 97.32% accuracy, 0.9971 sensitivity, 0.9598 specificity, and 0.9326 PPV. Moreover, the comparative experiments with respect to supervised learning and training strategies demonstrate that the proposed method is able to improve the diagnostic accuracy and robustness without exhaustive labeling. The proposed semi-supervised method, exploiting both supervised and unsupervised learning, facilitates an accurate and reliable diagnosis for COVID-19, leading to an improved patient care and management.